Hi Daniel,

You can try using VCFtools. It will give you global as well as pairwise Fst. It is easy and can compute fst from the vcf itself.

Else the R package 'hierfstat' should definitely help.

Hope this is of some help.

Regards,

Sorry, my question was not complete actually, I can get the global Fst with Hierfstat (Weir and Cockerham), what I cannot get, and need, are the CIs for that estimate. When using the boot.vc() function it gives back the error:

Error in boot.vc(all_loci_hierfstat) :

promise already under evaluation: recursive default argument reference or earlier problems?

I like the R package assigner (https://github.com/thierrygosselin/assigner). Provides a global estimate along with pairwise. Also does CIs.

Yeah, the problem with assigner (in my case at least) is that I fail to convert my genind (or genlight) to a "tidy genomic object"

I've had that problem too and usually had to brute force my way to fix it. Check the input data and then the genind: make sure the alleles and loci names are coded correctly (non-alphanumeric characters in loci names have been my problem in the past). Sometimes I have converted to a different format using radiator, then used the tidy function to generate a tidy genomic object (going from a VCF to a genind seems to create problems for these packages).

There is also the mmod package that calculates Fst and works with genind objects (haven't used it in a long time, though). StrataG, dartR, and SambaR are also packages that can calculate Fst and handle different data types, but I haven't used any of them.

If I recall correctly, the diveRsity package does via the bigDivPart() function.

Thank you Sean M. Murphy,

but unfortunately, that function is no longer in use, instead it is diffCalc, and it takes forever for a large dataset. Following a rule of thumb I would like to go through 1000 bootstraps. Perhaps if I had a more potent laptop (or complete permissions over it) I could make use of it, but not even indicating para = 4 gets the job done before one day.